<div dir="ltr">Your suggestion of rebuilding the driver fixed the issue of the UPS turning back during a power failure. Now the UPS shuts down the clients, then powers off appropriately. If power returns it turns back on also.<br><br>The only bug is if power returns before the UPS has turned the load off. In that case the UPS will remain off until it is manually turned on.<br><br>This might just be a limitation of this model, I don't know. But at least this behavior is not catastrophic like before. I suppose I can live with this edge case bug.<br><br>Thank you for your help again!<br><br>I don't know if my methods are perfect but just in case anyone has the same issue, here is the change I made:<br><br><a href="https://github.com/Laynezilla/nut/commit/ce3fcd0533e66b8be4adcf63ea0a63a7c45ba024">https://github.com/Laynezilla/nut/commit/ce3fcd0533e66b8be4adcf63ea0a63a7c45ba024</a><br><br>And here are the commands I used:<br><br>sudo apt update<br>sudo apt install devscripts git<br>mk-build-deps nut --install --root-cmd sudo --remove<br>sudo apt remove asciidoc<br>cd ~/<br>git clone <a href="https://github.com/Laynezilla/nut.git">https://github.com/Laynezilla/nut.git</a><br>cd ~/nut<br>./autogen.sh<br>./configure \<br><div style="margin-left:40px"> --build=`gcc -print-multiarch` \<br> --includedir=/usr/include \<br> --mandir=/usr/share/man \<br> --infodir=/usr/share/info \<br> --sysconfdir=/etc \<br> --localstatedir=/var \<br> --disable-silent-rules \<br> --libdir=/usr/lib/`gcc -print-multiarch` \<br> --runstatedir=/run \<br> --enable-maintainer-mode \<br> --disable-dependency-tracking \<br> --sysconfdir=/etc/nut \<br> --includedir=/usr/include \<br> --mandir=/usr/share/man \<br> --libdir=/lib/`gcc -print-multiarch` \<br> --with-ssl \<br> --with-nss \<br> --with-cgi \<br> --with-dev \<br> --enable-static \<br> --with-statepath=/run/nut \<br> --with-altpidpath=/run/nut \<br> --with-drvpath=/lib/nut \<br> --with-cgipath=/usr/lib/cgi-bin/nut \<br> --with-htmlpath=/usr/share/nut/www \<br> --with-pidpath=/run/nut \<br> --datadir=/usr/share/nut \<br> --with-pkgconfig-dir=/usr/lib/`gcc -print-multiarch`/pkgconfig \<br> --with-user=nut \<br> --with-group=nut \<br> --with-udev-dir=/lib/udev \<br> --with-systemdsystemunitdir=/lib/systemd/system \<br> --with-systemdshutdowndir=/lib/systemd/system-shutdown<br></div>make<br>make check VERBOSE=1<br>cd ~/nut/drivers<br>sudo upsdrvctl stop<br>sudo make install<br>sudo apt remove nut-build-deps<br>sudo apt remove devscripts<br>sudo apt autoremove<br>sudo upsdrvctl start</div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Sat, Jan 4, 2020 at 5:15 PM Ken Olum <<a href="mailto:kdo@cosmos.phy.tufts.edu">kdo@cosmos.phy.tufts.edu</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"> From: Layne Fowler <<a href="mailto:layne.fowler@gmail.com" target="_blank">layne.fowler@gmail.com</a>><br>
Date: Fri, 3 Jan 2020 16:22:10 -0500<br>
<br>
I see, so you kept the load.on.delay command but removed the check to see<br>
if it was handled. Did you do the same thing for the shutdown.stayoff logic?<br>
<br>
I didn't, because I don't use that functionality, though it does look as<br>
though it should be modified in just the same way.<br>
<br>
This might be a dumb question but did this require you to recompile the<br>
kernel? These types of changes are way beyond my depth.<br>
<br>
I rebuilt NUT (saying "make install" seems to have been sufficient). I<br>
didn't do anything involving my Linux kernel.<br>
<br>
Ken<br>
</blockquote></div>